Gabber

Gabber positions its voice generation as a cheaper alternative to ElevenLabs and Play.ht, though no specific plan prices are given here.

It is a real-time AI voice infrastructure for developers who build voice apps, agents and personas, and its main output is speech from Orpheus Voice TTS, which is described as low latency and emotionally expressive.

Beyond text-to-speech, Gabber covers the other pieces a voice app needs.

You can build agents with HD voice, text chat, memory and tool-calling, so the AI can talk, text, see, switch contexts, call tools and remember earlier conversation.

The company describes Gabber as the inference engine behind real-time AI apps and personas.

That makes Gabber a better fit if you are assembling a conversational product than if you only need a one-off voice clip.

It sits in the developer tools space rather than being a consumer voice generator, and the tags also include voice cloning.

Expect to wire it into your own app instead of using it as a standalone editor.